I think this is probably caused by UTF-8 not being specified everywhere
in the encoding chain.
The Roma interface doesn't seem to specify any encoding on its HTML
documents at all. It would be a good idea to add an HTML meta tag
specifying the encoding, because if no encoding is specified, my
understanding is that 8859-1 is the default. Not what you want, since it
can't handle cyrillic.
The form that's going to be submitted to the PHP handler should also
have its encoding specified using the accept-charset attribute. It
currently has no setting at all.
So I think the first changes to try would be to add this to the HTML head:
<meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
and to add this attribute to the <form> element:
There might be a hoop or two to jump through on the server side too --
not sure how that's working.
On 12-10-22 11:20 AM, Sebastian Rahtz wrote:
> Good gracious. That's a really dramatic effect. Its obviously an encoding problem (as ever),
> and the log gives some clues:
> [Mon Oct 22 19:00:09 2012] [error] [client 126.96.36.199] PHP Warning: DOMDocument::loadXML(): Entity 'ETH' not defined in Entity, line: 2 in /TEI/Roma/notam/notamHandler.php on line 39, referer: http://tei.oucs.ox.ac.uk/Roma/startroma.php?mode=changeElement&element=abbr&module=core
> [Mon Oct 22 19:00:09 2012] [error] [client 188.8.131.52] PHP Warning: DOMDocument::loadXML(): Entity 'deg' not defined in Entity, line: 2 in /TEI/Roma/notam/notamHandler.php on line 39, referer: http://tei.oucs.ox.ac.uk/Roma/startroma.php?mode=changeElement&element=abbr&module=core
> because, apparently, the PHP processor has deconstructed your Cyrillic into gibberish entities.
> How to fix this? Frankly, I haven't a clue! If anyone else reading this can help, I'd be very glad.
> Sebastian Rahtz
> Director (Research Support) of Academic IT Services
> University of Oxford IT Services
> 13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
University of Victoria Humanities Computing and Media Centre
([log in to unmask])